Reply to Amor et al

Microduplications on chromosome 14q11.2 including the FOXG1 gene have been reported in patients with developmental delay, cognitive impairment with speech delay, and epilepsy.^{1, 2} Such association has been confirmed subsequently in other patients.^{3, 4} However, Amor et al⁵ have found an ∼88 kb duplication at 14q12, encompassing the FOXG1 and C14orf23 genes in a father-son pair with isolated hemifacial microsomia. Neither the son nor the father exhibited mental retardation or epilepsy. They also identified an ∼3 Mb duplication of the 14q12 region, including FOXG1, in a child enrolled as a control subject in the CHOP CNV database⁶ and questioned the pathogenicity of FOXG1 duplication.

We believe it is important to highlight that the clinical phenotypes observed in the seven patients in the original description of the syndrome include a relatively wide spectrum of neurodevelopmental abnormalities, ranging from mild to severe intellectual disability and variable presence of epilepsy (in four out of the seven patients).¹ Thus, it is not surprising that subjects at the mildest end of the spectrum may present with few or no clinically evident manifestations of the disease.

Moreover, the duplicated copy of FOXG1 reported by Amor et al⁵ is small and may be devoid of its distant regulatory elements, which may explain the lack of neurocognitive phenotype. In support of this notion, two other FOX genes, FOXF1 and FOXL2, encoding for the evolutionarily conserved family of transcription factors with a central role in development have been shown recently to be upregulated by non-coding copy-number variants mapping over 250 kb 5′ from these genes.^{7, 8} Of note, FOXG1 expression is restricted to the brain, and thus more likely to be finely regulated by such distant enhancer(s) in a tissue-specific manner.

With regard to the individual in the CHOP CNV database with a FOXG1 duplication, we agree with the authors' suggestion that the CHOP subject with the duplication of FOXG1 may have not manifested yet the neurodevelopmental abnormalities.

On the basis of the well-established causative role of genomic deletions and point mutations of FOXG1 in determining a Rett-like phenotype^{9, 10} and the studies generated in animal models,¹¹ the evidence of FOXG1 as a dosage sensitive gene is compelling. Thus, we believe microduplications involving FOXG1 should not be considered of questionable pathogenicity but rather highly likely to be pathogenic, albeit associated with a wide spectrum of abnormalities, which is commonly observed with other microduplication syndromes.¹²

Drs Cheung and Stankiewicz are based in the Department of Molecular and Human Genetics at Baylor College of Medicine (BCM), which offers genetic laboratory testing, including use of arrays for genomic copy number analysis, and derives revenue from this activity.